Proteomics Based Analysis of the Nicotine Catabolism In

www.nature.com/scientificreports

OPEN Proteomics based analysis of the nicotine catabolism in Paenarthrobacter nicotinovorans Received: 11 February 2018 Accepted: 24 October 2018 pAO1 Published: xx xx xxxx Marius Mihăşan 1,2, Cornelia Babii1, Roshanak Aslebagh2, Devika Channaveerappa2, Emmalyn Dupree2 & Costel C. Darie2

Paenarthrobacter nicotinovorans is a nicotine-degrading microorganism that shows a promising biotechnological potential for the production of compounds with industrial and pharmaceutical importance. Its ability to use nicotine was linked to the presence of the catabolic megaplasmid pAO1. Although extensive work has been performed on the molecular biology of nicotine degradation in this bacterium, only half of the genes putatively involved have been experimentally linked to nicotine. In the current approach, we used nanoLC–MS/MS to identify a total of 801 proteins grouped in 511 non- redundant protein clusters when P. nicotinovorans was grown on citrate, nicotine and nicotine and citrate as the only carbon sources. The diferences in protein abundance showed that deamination is preferred when citrate is present. Several putative genes from the pAO1 megaplasmid have been shown to have a nicotine-dependent expression, including a hypothetical polyketide cyclase. We hypothesize that the enzyme would hydrolyze the N1-C6 bond from the pyridine ring with the formation of α-ketoglutaramate. Two chromosomally-encoded proteins, a malate dehydrogenase, and a D-3- phosphoglycerate dehydrogenase were shown to be strongly up-regulated when nicotine was the sole carbon source and could be related to the production the α-keto-glutarate. The data have been deposited to the ProteomeXchange with identifer PXD008756.

Nicotine is the main alkaloid produced by the tobacco plant as an anti-herbivore chemical. Due to its potent parasympathomimetic stimulant efect, nicotine plays a crucial role in smoking addiction and proved to be toxic in high concentrations to both animals and humans1. Nicotine can accumulate as both liquid and solid tobacco waste during the manufacturing of tobacco products. As it is water soluble, nicotine can easily contaminate the environment and endanger human health and aquatic life. Several bacterial strains have been shown to use nicotine as their sole carbon and nitrogen source for growth. Generally referred as Nicotine-Degrading Microorganisms (NDMs)2, these bacteria offer an eco-friendly method for removing nicotine from tobacco products and waste by means of bioremediation. The NDMs could also provide enzymes and pathways that can be engineered to convert nicotine and other pyridine and/ or pyrrolidine ring containing compounds into starting materials for the synthesis of products of industrial and pharmaceutical importance. Te development of technologies for converting nicotine waste into 6-Hydroxy- 3-succinoyl-pyridine3, 3-succinoyl-pyridine4 or 6-hydroxy-nicotine5 as well as the recent identification of 6-hydroxy-nicotine as a neuroprotective drug6 shows that NDMs nicotine pathways and enzymes have valuable biotechnological applications and could be used for the production of green chemicals. Paenarthrobacter nicotinovorans is a soil Gram-positive NDM that frst came into the limelight in the 1960’s under the name Arthrobacter oxidans7, re-classifed as Arthrobacter nicotinovorans8 in the 1990’s and received its current name in 20169. P. nicotinovorans ability to use nicotine as carbon and energy source is linked to the presence of a cluster of nicotine-inducible genes (nic- genes) placed on the catabolic megaplasmid pAO110. P. nicotinovorans makes use of the pyridine pathway to degrade nicotine. Briefy, nicotine catabolism starts with

1Biochemistry and Molecular Biology Laboratory, Department of Biology, Alexandru Ioan Cuza University of Iaşi, Iaşi, Romania. 2Biochemistry & Proteomics Group, Department of Chemistry & Biomolecular Science, Clarkson University, Potsdam, NY, USA. Correspondence and requests for materials should be addressed to M.M. (email: [email protected])

SCientifiC REPorts | (2018)8:16239 | DOI:10.1038/s41598-018-34687-y 1 www.nature.com/scientificreports/

hydroxylation of the pyridine ring and results in formation of γ-N-methylaminobutyrate (CH3-4-GABA) and 2,6-dihydroxy-pyridine (2,6-DHP). CH3-4-GABA is degraded to succinate and methylamine in what is called the lower nicotine pathway, the latter compound been shown to accumulate into the growth medium11. 2,6-DHP is hydroxylated to trihydroxypyridine (THP) which spontaneously dimerizes to 4,4′,5,5′-tetrahydroxy-3,3′-diaz adiphenoquinone-(2,2′) (Nicotine-blue, NB) giving the characteristic nicotine-blue color (for an overview see reference12, Fig. 1 and supplementary Fig. 1). Te remarkable work performed by the Decker and later the Brandsch group on the molecular biology and biochemistry of nicotine degradation in P. nicotinovorans13 lead to the experimentally establishment of functions for 19 out of the 40 genes making out the nic-gene cluster12. Te rest of the nic genes either have putative or no known functions or are may not be expressed at all. Tese genes might include some missing key points in nicotine metabolism such as major gene regulators or nicotine transporters. An experimental indication of the involvement of these putative genes in the nicotine metabolism is required. Te current shotgun proteomics approach attempts to fll this gap by identifying the proteins expressed by P. nicotinovorans in the presence of nicotine using nanoLC–MS/MS. A total of 801 proteins grouped in 511 non-redundant protein clusters were identifed when P. nicotinovorans was grown on citrate, nicotine and citrate, and nicotine as the only carbon source. Te diferences in protein expression patterns allowed us not only to fll in some of the blank spots in nicotine metabolism, but also to relate this plasmid-encoded catabolic pathway to the general chromosome-encoded pathways of the cell. Tis proteomics data can provide some valuable insight into the genetics and physiology of nicotine metabolism and serve as a basis for generating hypotheses for future attempts to genetically-engineer the catabolic pathway for increased bioremediation efciency or production of green chemicals. Results and Discussion Proteins identifcation overview. Te cell free lysates from Paenarthrobacter nicotinovorans pOA1 growing on citrate, nicotine and citrate and nicotine as the only carbon sources were frst separated by sodium dodecyl sulfate-polyacrylamine gel elctrophoresis (SDS-PAGE), then digested with trypsin and the resulting peptides analyzed by nanoLiquid Chromatography Tandem Mass Spectrometry (nanoLC-MS/MS). A bottom-up approach based on the MASCOT MS/MS database search algorithm was used to identify the proteins in each sample. Currently, there is no fnal genome of P. nicotinovorans pAO1 available, so for the database searches a custom database was used. Te database included the reference proteome of P. aurescens strain TC1, the complete proteome of pAO1 megaplasmid as well as the partial proteome derived from the draf genome of P. nicotinovorans strain Hce-1 (WGS sequence set BDDW01000000). Tis approach allowed us to identify a total of 801 proteins grouped in 511 non-redundant protein clusters. Te full list of proteins identifed in this study is presented in Supplementary Table 1. Te distribution of these clusters showing the overlap between the proteomes of P. aureus TC1 and P. nicotinovorans is depicted in Table 1. A total number of 368 non-redundant proteins were identifed based on genes present in both the P. aureus TC1 and the P. nicotinovorans making out the core-genome of the two species. Moreover, a number 115 non-redundant proteins hits were identifed based on genes that are only found in the P. nicotinovorans pangenome, 28 being encoded by the pAO1 megaplasmid. One unexpected fnding is that a number of 30 non-redundant proteins hits were identifed based on genes that are specifc for P. aureus TC1 genome. Tis can only be explained by the fact that the WGS sequence set BDDW01000000 is not the complete genes set for the P. nicotinovorans strain. Te overlap between the shared and unique proteins expressed in P. nicotinovorans pAO1 cells growing on dif- ferent carbon sources is summarized in the Venn diagram from Fig. 2. Disregarding whether it is used as the sole Carbon source, the presence of nicotine in the growth medium triggers several important changes in the protein expression patterns of P. nicotinovorans. Te magnitude and signifcance of these changes can be observed in the Volcano plots depicted in Fig. 3. Due to the specifc way the Scafold sofware calculates the fold change, the proteins not expressed in any of the compared conditions do not show up on the graphs. In the cells grown solely on nicotine, 46 out of 632 identifed proteins were dysregulated when compared to the citrate grown nicotine cells. When both nicotine and citrate were present as the carbon source, the number of dysregulated proteins was only 16 out of 618. Te full list of statistically signifcant dysregulated proteins in all growth conditions is presented in Supplementary Table 2.

Diferentially regulated plasmid-encoded proteins in P. nicotinovorans in the presence of nicotine. As expected, none of the proteins encoded by the pAO1 megaplasmid and known to be involved in nicotine metabolism could be identifed in the cell free extracts from the cells grown on citrate alone. Te presence of nicotine in the growth media alone or in combination with citrate leads to the expression of all the pAO1 enzymes known to be responsible for converting nicotine to γ-N-methylaminobutirate: nicotine dehydrogenase (NDH), 6-hydroxy-L-nicotine oxidase (6HLNO), 6-hydroxypseudooxynicotine dehydrogenase (KDH) and 2,6-dihydropseudooxynicotine hydrolase (DHPONH). Also, the enzymes responsible for the production of nicotine blue, the 2,6-dihydroxypyridine 3-monooxygenase (DHPH) and the NAD(P)H-Nicotine Blue Oxidoreductase (NBOR) were also identifed when nicotine was present in the growth medium (Table 2).

Deamination is preferred in the lower nicotine pathway when citrate can be used as a Carbon source. The lower nicotine pathway of P. nicotinovorans, consists of either a deamination or demethylation of 11 γ-N-methylaminobutyrate (CH3-4-GABA) resulted from the pyrrolidine ring cleavage. As shown in Fig. 1, in the deamination pathway, CH3-4-GABA is converted by MAO, a methylamine-forming 4-methylaminobutanoate oxidase (MABO2_PAENI) to succinic-semialdehyde. Te methylamine is excreted in the growth medium14, while the succinic-semialdehyde is further converted by the downstream enzyme succinic-semialdehide dehydrogenase SsaDH (SSDH_PAENI) to succinic acid and integrated into the tricarboxylic acid (TCA) cycle11. In the

SCientifiC REPorts | (2018)8:16239 | DOI:10.1038/s41598-018-34687-y 2 www.nature.com/scientificreports/

NDH6HLNO 6HDNO

KDH

DHPONH

DHPH MABO MAO

MAO

SsaDH PKC FolD NBOR PurU

NIT

Figure 1. Overview of the nicotine catabolic pathway of Paenarthrobacter nicotinovorans and the pAO1 encoded enzymes identifed in the current proteomics experiment. Blue text indicates enzymes shown to have a nicotine-dependent expression by the proteomics experiment. Magenta text indicates ORF’s with putative function related to nicotine and shown to have a nicotine-dependent expression by our proteomics experiment. Black text indicates enzymes that were not detected as expressed in our proteomics experiment. Light blue background indicates the lower nicotine pathway. CAPS indicate enzymes catalyzing the stepwise degradation of nicotine: NDH - nicotine dehydrogenase; 6HLNO - 6-hydroxy-L-nicotine oxidase; 6HDNO - 6-hydroxy-D-nicotine oxidase; KDH - ketone dehydrogenase; DHPONH - 2,6-dihydroxypseudooxynicotine hydrolase; DHPH – 2,6-dihydroxypyridine-3-hydroxylase NBOR – nicotine blue oxidoreductase; MABO - γ-N- methylaminobutyrate oxidase; FolD - methylene-tetrahydrofolate dehydrogenase/cyclohydrolase; PurU - formyltetrahydrofolate deformylase; MAO - monoamine-oxidase; SsaDH - succinic semialdehyde dehydrogenase; PKC – putative polyketide cyclase; NIT - ω-amidase. CAPS AND BOLD letters indicate the intermediates: 6HNic – 6-hydroxynicotine; 6-HMM – 6-hydroxy-methylmyosmine; 6-HPON – 6-hydroxy-pseudooxynicotine; 2,6-DHPON – 2,6-dihydroxypseudooxynicotine; 2,6-DHP – 6-dihydoxypyridine; CH3-4-GABA - γ-N- methylaminobutyrate; 2,3,6-THP - 2,3,6-trihydroxypyridine; NB -nicotine blue, 4,4′,5,5′-tetrahydroxy-3,3′- diazadiphenoquinone-(2,2′); CH2 TH4 - methylenetetrahydrofolate; GABA - γ-aminobutyric acid; SSA - succinic semialdehyde, alpha-KGA - a-keto-glutaramate; alpha-KG - a-keto-glutarate.

SCientifiC REPorts | (2018)8:16239 | DOI:10.1038/s41598-018-34687-y 3 www.nature.com/scientificreports/

Proteins Dimension Nicotine Total number Proteome (sequences) Citrate Nicotine and Citrate of proteins Paenarthrobacter aureus TC1 4565 332 306 218 397 Paenarthrobacter nicotinovorans 4722 377 369 268 482 Specifc to P. nicotinovorans — 66 85 62 115 pAO1 megaplasmid 175 1 21 27 28 Total number of non-redundant proteins — 398 391 279 511

Table 1. Distribution of the shared and specifc non-redundant proteins identifed in the cell free lysates of Paenarthrobacter nicotinovorans pAO1.

Figure 2. Venn diagram illustrating overlaps between the substrate specifc proteins identifed by LC-MS/MS analysis of Paenarthrobacter nicotinovorans pAO1 grown of diferent carbon sources.

demethylation pathway, CH3-4-GABA is converted by MABO, a formaldehyde-forming 4-methylaminobutanoate oxidase (MABO1_PAENI) in γ-aminobutyrate (GABA), methylenetetrahydrofolate (CH2-THF) and reduced 15 11 FADH2 . By using revers transcription PCR, Chiribau et al. showed the nicotine dependent expression of both mao, ssdh and mao genes and postulated that the both pathways must be active in-vivo. Using in vitro kinetic data, the authors also predicted a preferentially channeling of CH3-4-GABA to the demethylation pathway, despite the fact that methylamine was known to accumulate in the growth medium. Our MS/MS data showed an unexpected fnding: the lower nicotine pathway enzymes of P. nicotinovorans are diferently expressed based on the available C sources. Te MAO and SsaDH enzymes were identifed in both the nicotine, and nicotine and citrate containing media, thus explaining the previously reported accumulation of methylamine. MABO, the key enzyme in the demethylation pathway, was found only when nicotine alone was used as a carbon source. Moreover, folD and purU, two putative genes located on the same pAO1 encoded operon as the gene encoding ABO, are also expressed when nicotine is present, but citrate is absent. Te folD gene encodes a bifunctional protein (Q8GAI4_PAENI) putatively involved in the oxidation of CH2-THF to 10-formyltetrahydrofolate (10-CHO-THF). Te purU gene encodes a formyltetrahydrofolate deformylase (Q8GAI2_PAENI) that catalyzes the hydrolysis of 10-CHO-THF to formate and tetrahydrofolate (THF). Te sequence of enzymes MABO, FolD, PurU will thereby provide a way of transforming the methyl group from CH3-4-GABA in formaldehyde that can be assimilated by the RuMP cycle, serine pathway or directly by formaldehyde oxidation16. Apparently, when both nicotine and citrate can be used as a C source, P. nicotinovorans uses the deamination pathway to quickly extract the succinic acid from CH3-4-GABA and to use it in the TCA cycle. When carbon sources are scarce and only nicotine is available, the demethylation pathway is activated in P. nicotinovorans. Tis generates GABA that can be converted to succinic acid and used again in the TCA cycle, but has the advantage of generating one extra methyl group that can be used for the synthesis of sugars or amino acids.

SCientifiC REPorts | (2018)8:16239 | DOI:10.1038/s41598-018-34687-y 4 www.nature.com/scientificreports/

Figure 3. Volcano plots demonstrating the magnitude (x-axis) and signifcance (y-axis) of the protein comparisons between: (A) Nicotine versus Citrate, (B) Citrate versus Nicotine and Citrate and (C) Nicotine versus Nicotine and Citrate. Red dots represent signifcantly altered proteins (p < 0.001), whereas blue dots represent statistically insignifcant proteins. Te labels indicate the UniProKB IDs for the proteins listed in Supplementary Table 2.

Experimental evidence of a hypothetical polyketide cyclase involvement in nicotine metabolism. Another interesting finding is that a hypothetical polyketide cyclase (PKC, Q93NG4_PAENI) has a nicotine-dependent expression (Table 2). BLAST searches using the PKC amino acid sequence gave several domain hits: a specifc hit for pfam04199 (putative cyclases, E-value = 9.71e-30) and two non-specifc hits for COG1878

SCientifiC REPorts | (2018)8:16239 | DOI:10.1038/s41598-018-34687-y 5 www.nature.com/scientificreports/

Compared growth conditions Nicotine vs Citrate Nicotine and Citrate vs Citrate Nicotine vs Nicotine and Citrate Fold Fisher’s exact test Fold Fisher’s exact test Fold Fisher’s exact test Protein UniProt ID Change (p < 0.00245) Change (p < 0.00077) Change (p < 0.00249) 2,6-dihydropseudooxynicotine hydrolase, 1 DHPON_PAENI INF <0.00010 INF <0.00010 3 <0.00010 DHPONH 2 2,6-dihydroxypyridine 3-monooxygenase, DHPH DHPH_PAENI INF <0.00010 INF <0.00010 2.8 <0.00010 3 2-oxoglutaramate amidase, NIT NIT_PAENI INF <0.00010 INF <0.00010 2.3 0.011 4-methylaminobutanoate oxidase (formaldehyde- 4 MABO1_PAENI INF 0.00013 n.d — INF <0.00010 forming), MABO 4-methylaminobutanoate oxidase (methylamine- 5 MABO2_PAENI INF <0.00010 INF <0.00010 3.9 <0.00010 forming), MAO 6 6-hydroxy-L-nicotine oxidase 6-HLNO Q93NH4_PAENI INF <0.00010 INF 0.00029 27 <0.00010 6-hydroxypseudooxynicotine dehydrogenase 7 KDHA_PAENI INF <0.00010 INF <0.00010 4.2 <0.00010 complex subunit alpha, KDHA 8 Bifunctional protein FolD Q8GAI4_PAENI INF 0.00069 n.d — INF 0.0002 9 Formyltetrahydrofolate deformylase, PurU Q8GAI2_PAENI INF 0.0017 n.d — INF 0.0014 10 Hypothetical polyketide cyclase, PKC Q93NG4_PAENI INF <0.00010 INF <0.00010 1.4 0.15 NAD(P)H-Nicotine Blue (4,4′,5,5′-tetrahydroxy- 11 3,3′-diazadiphenoquinone-(2,2′)) Oxidoreductase, Q8GAJ2_PAENI INF <0.00010 INF <0.00010 1.4 0.23 NBOR 12 Nicotine dehydrogenase large subunit NDHL Q93NH5_PAENI INF <0.00010 INF 0.00031 3.1 0.0058 13 Nicotine dehydrogenase medium subunit, NDHM Q59127_PAENI INF <0.00010 INF <0.00010 2.7 0.0078 14 Nicotine dehydrogenase small subunit, NDHS Q59129_PAENI INF <0.00010 INF 0.00031 3.1 0.0058 ORF106, putative molybdenum transport ATPase, 15 Q93NG5_PAENI INF 0.0011 INF <0.00010 1.5 0.25 MODC ORF78, putative carbon monoxide dehydrogenase 16 Q93NG2_PAENI INF <0.00010 INF <0.00010 2 0.00017 subunit G, COXG ORF76, putative carbon monoxide dehydrogenase 17 Q93NG0_PAENI INF <0.00010 INF <0.00010 1.5 0.24 subunit D, COXD 18 Succinate-semialdehyde dehydrogenase SsaDH SSDH_PAENI INF <0.00010 INF 0.00023 7.4 0.00013 ORF92, hypothetical DUF948 domain-containing 19 Q8GAH6_PAENI INF 0.00032 n.d — INF 0.00014 protein

Table 2. Diferentially regulated pAO1-encoded proteins during the growth of Paenarthrobacter nicotinovorans on diferent carbon sources. aFold Change as calculated by Scafold v.4.8.2– weighted number of spectral counts for one protein in one condition vs weighted number of spectral counts in the second condition; a value is defned “signifcant” when it is either greater than or equal to 2.0 (protein is upregulated) or it is less than or equal to 0.5 (protein is down-regulated). INF - protein identifed in the frst condition, but not second, 0 - Protein identifed in the second condition, but not frst, n.d. - protein not detected in any of the compared conditions. Bold typeface indicates proteins that have been shown experimentally to be involved in nicotine catabolism.

(kynurenine formamidase, E-value = 7.58e-29) and TIGR03035 (arylformamidase, E-value = 6.14e-05). Te bacterial kynurenine formamidase are metal-dependent hydrolases17 acting on carbon-nitrogen bonds other than peptide bonds. As the pkc is apparently placed in the same operon with kdhL and dhph encoding the key enzymes for the production of 2,3,6-THP (Fig. 3), it is highly tempting to hypothesize that PKC would hydrolyze the N1-C6 bond in 2,3,6-THP with the formation of α-ketoglutaramate. An indirect indication that this might be the case is the fact that the 2-oxoglutaramate amidase (NIT, NIT_PAENI) is also expressed only when nicotine is present in the medium (Table 2). Te NIT activity converting α-keto-glutaramate to α-keto-glutarate was experimentally shown and it was postulated that the enzyme connects the nicotine pathway to the citric acid cycle and makes the pyridine nitrogen available for assimilation via a glutamate dehydrogenase18. Te 2-Hydroxypyridine degradation pathway in Rhodococcus rhodochrous strain PY11 also has THP as an intermedi- ate connects and to the citric acid cycle in a very similar manner19. It makes use of HpoH, a THP hydroxylase and HpoI, a 2-ketoglutaramate amidase. HpoH of R. rhodochrous PY11 and PKC of P. nicotinovorans pAO1 share a level of 77% sequence similarity and the two amidases share 47% sequence similarity. Diferentially regulated chromosomally-encoded proteins in P. nicotinovorans in the presence of nicotine. A striking fold change related to nicotine was recorded for two chromosomally-encoded enzymes: a malate dehydrogenase (A0A175JBL3_PAENI) and a D-3-phosphoglycerate dehydrogenase (PHGDH, A0A175J283_PAENI). In both cases, the enzymes are up-regulated when nicotine is the sole carbon source present in the medium but are not signifcantly dysregulated when both nicotine and citrate are present. In addition to catalyzing oxidation of 3-phosphoglycerate, PHGDH has been shown to catalyze the NADH-dependent reduction of α-keto-glutarate20 and its high expression would ft very well with the above mentioned expression of NIT and production of α-keto-glutarate from the pyridine ring of nicotine. Te malate dehydrogenase is involved in the oxidation of malate to oxaloacetate in many metabolic pathways, including the citric acid cycle. Te high expression levels of these two enzymes must be related to the cells efort to concentrate all the available carbon

SCientifiC REPorts | (2018)8:16239 | DOI:10.1038/s41598-018-34687-y 6 www.nature.com/scientificreports/

derived from nicotine metabolism into the major metabolic pathways of the cell required for growth. When citrate is abundant in the medium (alone or in combination with nicotine), such an efort is not required and the two enzymes return to a more basal expression. Another general metabolism enzyme that can be linked to the presence of nicotine in the growth medium is a catalase (A0A175J0M2_PAENI). Te enzyme has approx. 2.5-fold change when nicotine is present in the growth medium, disregarding weather citrate is present or not. Interestingly, the expression levels on nicotine only and nicotine and citrate media are the same. Te production of nicotine-blue pigment involves a spontaneous oxida- − tion reaction accompanied by the release of O2 . It was postulated that this oxidation reaction may represent a selective advantage for P. nicotinovorans bacteria in competition with other soil community bacteria sensitive to oxygen radicals21 and that the NAD(P)H-Nicotine Blue Oxidoreductase (NBOR) enzyme might prevent the formation of these radicals inside the cell. Our MS data indicate that the mechanism by which P. nicotinovorans itself appears not to be afected by the generation of these radicals during growth on nicotine is the increased expression of catalase and that the role of the NBOR must be reconsidered. Considering the MS/MS data showing the nicotine dependent expression of PKC and NIT, we can assume that Nicotine-blue reduction reaction catalyzed by the NBOR is actually preparing the pyridine ring to be cleaved by PKC. Te product of the cleavage reaction is then converted to α-keto-glutarate by the plasmid –encoded amidase NIT and integrated into the general metabolism by the chromosomally-encoded PHGDH. In conclusion, we used nanoLC–MS/MS to identify a total of 801 proteins grouped in 511 non-redundant protein clusters when P. nicotinovorans was grown on diferent C-sources. Te diferences in protein abundance showed that deamination is preferred in the lower nicotine pathway when citrate is present in the medium. A hypothetical polyketide cyclase was shown to have a nicotine-dependent expression and we hypothesize that the enzyme would hydrolyze the N1-C6 bond from the pyridine ring with the formation of α-ketoglutaramate. Two chromosomally-encoded proteins, a malate dehydrogenase, and a D-3-phosphoglycerate dehydrogenase were shown to be strongly up-regulated when nicotine was the sole carbon source and could be related to the production the α-keto-glutarate. Te proteomics data have been deposited to the ProteomeXchange with identifer PXD008756 and might serve as a basis for generating hypotheses for future attempts to genetically-engineer the catabolic pathway for increased bioremediation efciency or production of green chemicals. Materials and Methods Materials. All materials were purchased from Sigma-Aldrich (St. Louis, MO, US) if not stated otherwise. DNase 1 and RNase A were from Roche (Basel, Switzerland). HPLC grade water and Acetonitrile were from Fisher Chemical (Pittsburgh, PA, USA). LC-MS grade formic acid was from Fluka (Buchs, Switzerland) and Iodoacetamide from Calbiochem.

Bacterial strain and growing conditions. Paenarthrobacter nicotinovorans pAO1 (DSM 420 Deutsche Sammlung von Mikroorganismen und Zellkulturen, Braunschweig, Germany) was grown on 34 mM Na2HPO4-22 −1 mM KH2PO4 bufer, pH 7.0, 0.2% (NH4)2SO4, supplemented with 5% mineral solution, 0.1 mg ml biotin and 35 ug ml−1 kanamycin. As the carbon source 0.05% nicotine, 0.2% sodium citrate or the combination of the two were used. Bacterial cultures were incubated at 28 °C on a rotary shaker (Model 3013, GFL, Burgwedel, Germany) at 180 rpm for 10 hours until nicotine was depleted from the medium and the specifc nicotine-blue color is formed.

Cell-lysis and sample preparation. Bacteria were collected by centrifugation at 4,500 × g, 20 min from cultures grown in the presence or absence of nicotine; the bacterial pellets were washed twice in 10 mM TRIS/HCl pH 7.4 for the removal of the nicotine-blue dye. Cells were lysed according to the protocol of Vandera et al.22, with slight modifcations. 0.5 g of cells were resuspended in 5 ml 40 mM Tris–HCl, pH 7.6, 0.03% PMSF (w/v), 0.005% chloramphenicol (w/v), 18% sucrose (w/v) in the presence of 50 mg ml−1 lysozyme and incubated for 60 min at 37 °C. Afer centrifugation, cell pellets were washed with 10 mM Tris–HCl, pH 7.6 and resuspended in 5 ml lysis bufer containing 40 mM Tris–HCl, pH 7.6, 0.3% SDS (w/v), 60 mM DTT, 10 μg ml−1 DNase I and 10 μg ml-1 RNase A. Samples were incubated for 30 min at 95 °C with intervals of vigorous vortexing every 5 min and then centrifuged on an Eppendorf 5417R centrifuge at 14,000 g and 4 °C for 30 minutes. Te cell lysates were quickly placed on ice and stored at −20 °C until further processing. Protein concentrations were determined by the BCA method, using an assay kit (Sigma Aldrich, Germany) with bovine serum albumin (BSA) as a standard. One hundred μg of total proteins from the cell free lysates were separated on a home-made 9–16% SDS-PAGE maxi (20 × 20 cm) gradient gels and then stained by Commasie Brilliant Blue R250. Each biological sample was loaded in triplicate. Te gel lanes for diferent biological samples were divided into 20 gel pieces and then sub- jected to in-gel digestion using trypsin, as described previously23. Briefy, each gel piece was washed under moder- ate shaking at room temperature (RT) in HPLC grade water for 60 min and then in 50% (v/v) acetonitrile (ACN)/ HPLC grade water containing 50 mM ammonium bicarbonate (ABC) for 60 min. Afer a dehydration step with 100% ACN for 60 min, the gel pieces were dried under Speed Vac and the Cys residues were further reduced and alkylated using with 10 mM dithiothreitol (DTT) in 25 mM ABC for 60 min. at 60 °C and with 100 mM iodoacetamide in 25 mM ABC for 60 min. in the dark. Gel pieces were dehydrated and dried again. Te digestion was performed by rehydrating the gel pieces in 200 μl of a trypsin solution (10 ng/μl) and incubating overnight at 37 °C. Afer incubation, peptide extraction was carried out with 5% formic acid (FA) in 50/50 (v/v) 50 mM ABC/ ACN and with 5% FA in ACN (60 min. each). Extracted peptides were dried and then cleaned by reversed-phase chromatography using C18 ZipTips (EMD Millipore, Billerica, MA).

Nanoliquid chromatography tandem mass spectrometry (nanoLC-MS/MS). Te resulting peptide mixture was analysed by reversed phase nanoLiquid Chromatography Tandem Mass Spectrometry (nanoLC-MS/ MS) using a NanoAcquity UPLC (Waters, Milford, MA, USA) coupled to a Q-TOF Xevo G2 MS (Waters),

SCientifiC REPorts | (2018)8:16239 | DOI:10.1038/s41598-018-34687-y 7 www.nature.com/scientificreports/

according to published procedures24,25. Te peptides were loaded onto a 100 μm × 10 mm NanoAquity BEH130 C18 1.7 μm UPLC column (Waters) and eluted over a 180-min. gradient at a fow rate of 400 nL/min as follows: 1–45% organic solvent B (ACN containing 0.1% FA) over 1–120 min., 45–85% B (120–140 min.), constant 85% B (140–160 min.), 85%–2% B (160–165 min.) and then return to the initial conditions of 1% B (165–180 min.). The aqueous solvent A was 0.1% FA in HPLC water. The column was coupled to a Picotip Emitter Silicatip nano-electrospray needle (New Objective, Woburn, MA, USA). MS data acquisition involved survey 0.5 sec. MS scans (m/z range 350–2000) and automatic data dependent analysis (DDA) of the top six ions with the highest intensity, with the charge of 2+, 3+ or 4+. Te MS/MS (recorded over m/z of 50–2000) was triggered when the MS signal intensity exceeded 500 counts/sec. In survey MS scans, the six most intense peaks were selected for collision-induced dissociation (CID) and fragmented until the total MS/MS ion counts reached 6000 or for up to 1.1 sec. each. Te entire procedure used was previously described24. Ion intensity was adjusted using Leucine enkephalin (2 ng ul−1) in direct infusion ESI-MS at 0.5 μL/min. Calibration was performed for both precursor and product ions using direct infusion ESI-MS at 0.5 μL/min of 1 pmole GluFib (Glu1-Fibrinopeptide B) standard peptide with the sequence EGVNDNEEGFFSAR and the theoretical monoisotopic (2+) peak with m/z of 785.84.

Data analysis. Te raw fles were analyzed using ProteinLynx Global Server v.2.4 (Waters, Milford, MA, USA) and Mascot v.2.5.1 (Matrix Science, London, UK). Peak list fles were generated using the following param- eters: background subtraction of polynomial order fve adaptive with a threshold of 30%, two smoothings with a window of three channels in Savitzky-Golay mode and centroid calculation of top 80% of peaks based on a minimum peak width of four channels at half height. Database searches were performed using a customized database containing the complete reference proteome of Paenarthrobacter aurescens strain TC1 (UP000000637)26, the proteome of Paenarthrobacter nicotinovorans strain Hce-1 (UP000078426)27 and the complete proteome of pAO1 megaplasmid (GenBank: AJ507836.1)10. Additionally, Mascot was set up to search for contaminants in the com- mon Repository of Adventitious Proteins database (2012.01.01; the Global Proteome Machine). Search parame- ters were strict trypsin specifcity allowing for up to three missed cleavages sites, fragment ion mass tolerance of 1.30 Da and a parent ion tolerance of 0.8 Da. Oxidation of methionine was selected as variable modifcation and carbamidomethylation of Cys was specifed as a fxed modifcation. False positive levels were estimated via decoy database method with a reverse database appended at the end of the forward database28. Scafold (v.4.8.2, Proteome Sofware Inc., Portland, OR, USA) was used to assem- ble the resulting database hits corresponding for the same biological data into one analysis using the MudPIT (Multidimensional Protein Identifcation Technology)29 and to validate MS/MS based peptide and protein identifcations. Te protein and peptide false discovery rate (FDR) was set to 1%. Protein identifcations were accepted if the protein contained at least 2 identifed peptides. Protein probabilities were assigned by the Protein Prophet algorithm30. Proteins that contained similar peptides and could not be diferentiated based on MS/MS analysis alone were grouped to satisfy the principles of parsimony. Proteins sharing signifcant peptide evidence were grouped into clusters. All hits from the contaminants database were manually fltered out. Proteins were anno- tated with GO terms from Gene Ontology Consortium31. Label-free relative quantifcation was performed using Scafold (v.4.8.2, Proteome Sofware Inc., Portland, OR, USA) using weighted spectra and outputted as fold change. Signifcance was assessed using the Fisher’s Exact test and the p-values were further adjusted using the Benjamini and Hochberg correction to account for multiple comparisons30. An adjusted p-value of less than 0.05 was used to select proteins that were diferentially expressed between compared groups. Te mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium (http://proteomecentral.proteomexchange.org)32 via the MassIVE partner repository33 with the dataset identifer PXD008756. Availability of Materials and Data Te mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium (http://proteomecentral.proteomexchange.org) via the MassIVE partner repository with the dataset identifer PXD008756. References 1. Mayer, B. How much nicotine kills a human? Tracing back the generally accepted lethal dose to dubious self-experiments in the nineteenth century. Arch. Toxicol. 88, 5–7 (2014). 2. Liu, J. et al. Nicotine-degrading microorganisms and their potential applications. Appl. Microbiol. Biotechnol. 99, 3775–3785 (2015). 3. Wang, S. N. et al. ‘Green’ route to 6-hydroxy-3-succinoyl-pyridine from (S)-nicotine of tobacco waste by whole cells of a Pseudomonas sp. Environ. Sci. Technol. 39, 6877–80 (2005). 4. Yu, H., Tang, H. & Xu, P. Green strategy from waste to value-added-chemical production: efcient biosynthesis of 6-hydroxy-3- succinoyl-pyridine by an engineered biocatalyst. Sci. Rep. 4, 5397 (2014). 5. Yu, W. et al. Green route to synthesis of valuable chemical 6-hydroxynicotine from nicotine in tobacco wastes using genetically engineered Agrobacterium tumefaciens S33. Biotechnol. Biofuels 10, 288 (2017). 6. Hritcu, L. et al. Nicotine versus 6-hydroxy-l-nicotine against chlorisondamine induced memory impairment and oxidative stress in the rat hippocampus. Biomed. Pharmacother. 86, 102–108 (2017). 7. Decker, K., Eberwein, H., Gries, F. & Bruehmueller, M. [On the degradation of nicotine by bacterial enzymes]. Hoppe-Seyler’s Zeitschrif fur Physiol. Chemie 319, 279–82 (1960). 8. Kodama, Y., Yamamoto, H., Amano, N. & Amachi, T. Reclassifcation of two strains of Arthrobacter oxydans and proposal of Arthrobacter nicotinovorans sp. nov. Int. J. Syst. Bacteriol. 42, 234–9 (1992). 9. Busse, H.-J. Review of the taxonomy of the genus Arthrobacter, emendation of the genus Arthrobacter sensu lato, proposal to reclassify selected species of the genus Arthrobacter in the novel genera Glutamicibacter gen. nov., Paeniglutamicibacter gen. nov., Pseudogluta. Int. J. Syst. Evol. Microbiol. 66, 9–37 (2016). 10. Igloi, G. L. & Brandsch, R. Sequence of the 165-kilobase catabolic plasmid pAO1 from Arthrobacter nicotinovorans and identifcation of a pAO1-dependent nicotine uptake system. J Bacteriol 185, 1976–1986 (2003). 11. Chiribau, C. B. et al. Final steps in the catabolism of nicotine - Deamination versus demethylation of gamma-N- methylaminobutyrate. Febs J. 273, 1528–1536 (2006).

SCientifiC REPorts | (2018)8:16239 | DOI:10.1038/s41598-018-34687-y 8 www.nature.com/scientificreports/

12. Ganas, P., Igloi, G. L. & Brandsch, R. Microbial Megaplasmids. Microbial Megaplasmids 11, (Springer Berlin Heidelberg 2009). 13. Brandsch, R. Microbiology and biochemistry of nicotine degradation. Appl Microbiol Biotechnol 69, 493–498 (2006). 14. Ganas, P., Mihasan, M., Igloi, G. L. & Brandsch, R. A two-component small multidrug resistance pump functions as a metabolic valve during nicotine catabolism by Arthrobacter nicotinovorans. Microbiology 153, 1546–55 (2007). 15. Chiribau, C. B., Sandu, C., Fraaije, M., Schiltz, E. & Brandsch, R. A novel gamma-N-methylaminobutyrate demethylating oxidase involved in catabolism of the tobacco alkaloid nicotine by Arthrobacter nicotinovorans pAO1. Eur. J. Biochem. 271, 4677–84 (2004). 16. Caspi, R. et al. Te MetaCyc database of metabolic pathways and enzymes and the BioCyc collection of pathway/genome databases. Nucleic Acids Res. 44, D471–D480 (2016). 17. Kurnasov, O. et al. Aerobic tryptophan degradation pathway in bacteria: novel kynurenine formamidase. FEMS Microbiol. Lett. 227, 219–227 (2003). 18. Cobzaru, C., Ganas, P., Mihasan, M., Schleberger, P. & Brandsch, R. Homologous gene clusters of nicotine catabolism, including a new ω-amidase for α-ketoglutaramate, in species of thre genera of Gram-positive bacteria. Res. Microbiol. 162, 285–91 (2011). 19. Vaitekunas, J., Gasparaviciute, R., Rutkiene, R., Tauraite, D. & Meskys, R. A 2-Hydroxypyridine Catabolism Pathway in Rhodococcus rhodochrous Strain PY11. Appl. Environ. Microbiol. 82, 1264–73 (2016). 20. Fan, J. et al. Human Phosphoglycerate Dehydrogenase Produces the Oncometabolite d -2-Hydroxyglutarate. ACS Chem. Biol. 10, 510–516 (2015). 21. Mihasan, M., Chiribau, C. B., Friedrich, T., Artenie, V. & Brandsch, R. An NAD(P)H-nicotine blue oxidoreductase is part of the nicotine regulon and may protect Arthrobacter nicotinovorans from oxidative stress during nicotine catabolism. Appl. Environ. Microbiol. 73, 2479–2485 (2007). 22. Vandera, E., Samiotaki, M., Parapouli, M., Panayotou, G. & Koukkou, A. I. Comparative proteomic analysis of Arthrobacter phenanthrenivorans Sphe3 on phenanthrene, phthalate and glucose. J. Proteomics 113, 73–89 (2015). 23. Ngounou Wetie, A. G. et al. Comparative two-dimensional polyacrylamide gel electrophoresis of the salivary proteome of children with autism spectrum disorder. J. Cell. Mol. Med. 19, 2664–78 (2015). 24. Channaveerappa, D. et al. Atrial electrophysiological and molecular remodelling induced by obstructive sleep apnoea. J. Cell. Mol. Med. 21, 2223–35 (2017). 25. Sokolowska, I., Dorobantu, C., Woods, A. G., Macovei, A., Branza-Nichita, N. & Darie, C. C. Proteomic analysis of plasma membranes isolated from undiferentiated and diferentiated HepaRG cells. Proteome Science 10(1), 47 (2012). 26. Mongodin, E. F. et al. Secrets of soil survival revealed by the genome sequence of Arthrobacter aurescens TC1. PLoS Genet 2, e214 (2006). 27. Liang, H. et al. Paenarthrobacter nicotinovorans strain Hce-1 geneom sequencing. Submitt. to EMBL/GenBank/DDBJ databases (2016). 28. Wang, G., Wu, W. W., Zhang, Z., Masilamani, S. & Shen, R.-F. Decoy Methods for Assessing False Positives and False Discovery Rates in Shotgun Proteomics. Anal. Chem. 81, 146–159 (2009). 29. Delahunty, C. M. & Yates, J. R. MudPIT: multidimensional protein identifcation technology. Biotechniques 43, 563, 565, 567 passim (2007). 30. Nesvizhskii, A. I., Keller, A., Kolker, E. & Aebersold, R. A statistical model for identifying proteins by tandem mass spectrometry. Anal. Chem. 75, 4646–58 (2003). 31. Ashburner, M. et al. Gene Ontology: tool for the unifcation of biology. Nat. Genet. 25, 25–29 (2000). 32. Deutsch, E. W. et al. The ProteomeXchange consortium in 2017: supporting the cultural change in proteomics public data deposition. Nucleic Acids Res. 45, D1100–D1106 (2017). 33. Perez-Riverol, Y., Alpi, E., Wang, R., Hermjakob, H. & Vizcaíno, J. A. Making proteomics data accessible and reusable: current state of proteomics databases and repositories. Proteomics 15, 930–49 (2015). Acknowledgements Tis work was supported by was supported by a grant from the Romanian National Authority for Scientifc Research and Innovation, CNCS-UEFISCDI PN-III-P2-2.1-PED-2016–0177. MM was supported by the Fulbright Senior Postdoctoral Fellowship awarded by the Romania-USA Fulbright Commission to MM (guest) and CCD (host). Author Contributions M.M. and C.C.D. devised the project, conceived and planned the experiments; M.M. and C.B. carried out the experiments and prepared the samples; R.A., D.C., E.D. performed the MS/MS determinations and acquired the data; M.M. and C.C.D. contributed to the interpretation of the results; M.M. took the lead in writing the manuscript; all authors reviewed and proof read the manuscript; C.C.D. supervised the project. Additional Information Supplementary information accompanies this paper at https://doi.org/10.1038/s41598-018-34687-y. Competing Interests: Te authors declare no competing interests. Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional afliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre- ative Commons license, and indicate if changes were made. Te images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

SCientifiC REPorts | (2018)8:16239 | DOI:10.1038/s41598-018-34687-y 9